The Journal of Organic Chemisty
2746 MANNING AND COLEMAN
Formation of Pyrazoles from 3,3-Disubstituted 2,P-Pentanediones and 2-Hydroxyethyl Hydrazines. Coproduction of 3,3a,5,6-Tetrahydropyrazolo[3,2-b]oxazoles DAVIDT. MANNINGAND HAROLD A. COLEMAN Research and Development Department, Union Carbide Corporation, Chemicals and Plastics, South Charbston, West Virginia 26303 Received January 16, 1969 Condensation of various 3,3-disubstituted 2,4pentanediones with 2-hydroxyethylhydrasines affords 3,3a,5,6 tetrahydropyrasolo [3,2-b]oxazoles resulting from an intramolecular alcohol-enamine reaction of the initially formed 5-methylene-2-pyrasolines. Nuclear magnetic resonance studies of these nonplanar heterobicycleswhich have been isolated indicate that phenyl, when attached to C-5, generally occupies a single preferred configuration. When one or both of the starting diketone substituents is allylic, a rearrangement, evidently of the Claisen-Cope type, occurs producing isomeric pyrasoles as coproducts. Tetrahydropyrasolo [3,2b]oxazoles bearing one or more allyl groups at C-3 undergo isomerization to their pyrasole coproducts under thermal conditions (ca. 200') or upon simple refluxing in ethanol solution. This latter transformation of 3-allyltetrahydropyrasolo[3,2-b]oxaaoles suggests that they are kinetically favored products which give rise to their thermodynamically stable pyrazole isomers during the course of the condensation reaction.
Earlier we noted that 3,3-disubstituted 2,4-pentanediones having allylic or propargylic groups a t C-3 react with monosubstituted hydrazines forming rearranged pyrazoles via a Claisen-Cope type of rearrangement. ' By means of an exceptionally facile propargylic rearrangement, the reaction enables preparation, in good yields, of pyrazoles having C-5 allenic substitution. The aim of obtaining such unsaturated pyrazoles with additional pharmacophoric 8-phenylethyl substitution suggested a-(hydrazinomethyl) benzyl alcohol (1) as an appropriate carbonyl reactant.
groups a t the pyrazoline C-4 of intermediates such as 3 rearrange more rapidly than do propargylic groups. SCHEME I
~
C
CHJHNHZ H O H
+
I I
F"
4X,CH,C=CH I
N
~ H Z
OH
(2) with 1 was performed in refluxing ethanol under conditions typical' for formation of 4 by propargylic re-
arrangement (Scheme I). The product, obtained in 59% yield, displayed none of the expected spectroscopic features of 4, however, but was a nonhydroxylic isomer with strong C-0--C (9.95 p ) absorption. The nmr spectrum showed two methyl groups and three AMX protons as pairs of doublets at 6 3.07, 3.99, and 4.65. These data, along with mass spectral studies, led to formulation of the new product as 3,3aj5,6-tetrahydro2,3a-dimethyl- 5 -phenyl - 3,3 - di(2 - propyny1)pyrazolo[3,2-b]oxazole (5) representing a heterobicyclic system described only recently by Gillis and Weinkam. Formation of 5 'may be viewed as an intramolecular nucleophilic interception of the Claisen-Cope precursor 3 by an alcohol-enamine addition related to the oxidative cyclization of 2-pyrrolidinoethanols reported by Leonard and M ~ s l t e r . ~We had observed that allylic (1) D.T. Manning, H. A. Coleman, and R. A. Langdale-Smith [ J . Org. Chem., 88, 4413 (1968)l (describe the general background of and lead references pertinent to this new variant of the Claisen-Cope rearrangement. (2) B.T.Gillis and R. Weinkam, ibid., 82, 3321 (1967). (3) N. J. Leonard and W. K. Musker, J . Amer. Chem. Soc., 82, 6148 (1980). By analogy with the oxidative cyclination of these authors the protonated iminium species 88 may be Considered to be the precuraor of the bicyclic system.
CH&=CH
CHZbCH 3.
-n,o __t
2
Results and Discussion Treatment of 3,3-di(2-propynyl)-2,4-pentanedione
(CH3C0)2C(CH2CeCH)2
3
N/
,CH2C=CH
I$
5
(iH3 N
A
bH
'CH~CH=C=CH~ 4
Thus it was not unexpected that in the reaction of 3,3diallyl-2,4-pentanedione (6) with 1 rearrangement competed with enamine cyclization giving both the pyrazole 10 and the tetrahydropyrazolo [3,2-b]oxazole 9 in crude yields of 54 and 24%, respectively (Scheme 11). The relatively nonpolar 9 was readily separated from 10 by column chromatography. Condensation of 1 with 6 to give the carbinol 7 is a reasonable first step and the ability to form enamines such as 8 under these reaction conditions has been demonstrated.l Compound 9 may then arise via paths A or B. The ability of 9 to serve as the precursor of 10
Vol. 34, No. 9, September 1969
3,3a,5,6-TETRAHYDROPYRALOLO[3,2-b]OXAZOLES 2747 SCHEME I1
1
+
(CHJO)&(CH2CH=CH)Z
--j
6
H 7
-Y
1*
FH3 /
8
h
’iH3
’iHS
was shown by refluxing 9 with ethanol under simulated reaction conditions. An approximately 85% conversion of 9 into 10 occurred suggesting that 3-allylic tetrahydropyrazolo [3,2-b]oxazolesare kinetically controlled products which are converted into their thermodynamically stable pyrazole isomers during the reaction by paths C and D. Similarly, synthesis of the 3-allyl-3propargyl derivative 11 resulted in coproduction of 12 along with a small amount of the product (13) of propargylic rearrangement (Scheme 111). Attempted vacuum distillation of 11 also gave 12, containing a trace of 13, in high yield. Tetrahydropyrazolo [3,2-b]oxazoles were the only products observed upon reaction of hydroxyethylhydrazines with 3,3dimethyl-2,4-pentanedione. Inferences from Nmr Spectral Data.-The rigid tetrahydropyrazolo [3,2-b]oxazole system b butterfly shaped with magnetically nonequivalent “concave” and “convex” surfaces. This accounts for the 6 2.55-2.82 resonance range of the propargyl methylenes of 5 (at C-3) and for the appearance of two C-3 methyl singlets for the tetramethyl derivative 14. With two
SCHEME I11
CH,C=CH --c+
c6H5
H 11
CH,C=CHz H CHZCH*CH=CHZ
c6H5
+
H 12
N A CH2CH=CHz
CH~C~CH 5
’CH, 14
centers of asymmetry 5 can exist, theoretically, as two stereoisomers. The appearance of both the C-2 and C-3a methyl resonances of 5 as sharp singlets indicates that either (a) a mixture of cis and trans isomers exists but this variation in phenyl position is without effect on the chemical shift of the methyl protons, or (b) the C-5 phenyl group exists in a single preferred configuration.
2748 MANNINQ AND COLEMAN
The Journal of Organic Chemistry
Consideration of the above-mentioned AMX patterns of the 5-phenylte~;rahydropyrazolo[3,2-b ]oxazoles enables a choice between these alternatives. While we are not able to give configurational assignments to H a , H,, H,, and phenyl, the appearance of three sharp pairs of doublets, with no second A m sets apparent, indicates that the phen.yl group occupies one preferred configuration and 5 is therefore a single stereoisomer. In the case of 11 the unlike allyl and propargyl groups at C-3 form a ca. 50:50 cis-trans mixture with the C-2 and C-3a methyls consequently showing pairs of singlets of approximately equal area. Unlike phenyl substitution, 15, having an aminomethyl group at C-5, was isolated as an isomer mixture of ca. 2 : 1 as estimated from the in-
Heywood and Phillips.l0 The dihydrochloride, mp 212-216' dec, was analyzed. Anal. Calcd for CgH18N20C12: N, 10.50. Found: N, 10.32. a-(Hydrazinomethyl)-4-methylpiperazineI-ethanol.-1-Glycidyl-4-methylpiperazine (78.1 g, 0.5 mol) and anhydrous hydrazine (165.0 g, 5.0 mol) were allowed to react under conditions similar to those employed in the preparation of 1. Distillation and redistillation gave 55.8 g of product, bp 132-135' (0.35 mm), whose nmr spectrum was confirmatory but showed the presence of some remaining butanol solvent. Anal. Calcd for CsH20N40: N, 29.76. Found: N, 28.13. 3,3a,5,6-Tetrahydro-2,3a-dimethyl- 5 -phenyl- 3,3 - di(2-propynyl)pyrazolo[3,2-b]oxazole@).-To a mixture of 1 (7.6 g, 0.05 mol) and 2 (8.8 g, 0.05 mol) in 200 ml of ethanol was added 15 ml of 5% acetic acid and the resulting mixture was then refluxed for a period of 3.5 hr. Evaporation of volatiles gave an oil which crystallized on standing. The crude product was dissolved in methanol and precipitated with water giving 7.05 g of cream-colored crystals, mp 99-101'. An impure fraction (1.55 g, mp 72-95') brought the total yield to 58.970: ir (KBr) 3.04 ( C k C H ) , 6.16 (C=N), 7.24 (CCHs), 9.95 (COC), 13.25 and 14.32 p (monosubstituted phenyl); mass spectrum [70 eV (49')] m/e 292 (parent), 186 (loss of CeH5CHCH2 ion), 147 (ion a), 104 (CeHsCHCH2 ion), 91 (benzyl ion), 77 (phenyl ion),
15
CH2N
tegrals of the two C-3a methyl signals. Here, the usual AMX pattern was obscured by the apparent presence of isomers and by additional coupling with the methylene protons attached a t C-5. Surprisingly, the apparent isomer mixture melted relatively sharply at 55.8-58". Two additional recrystallizations changed the isomer ratio, indicated by the C-3a methyl integrals, to 3-5: 1, but with only a small change in melting point. Despite the nmr evidence of stereoisomerism, both 11 and 15 behaved as single compounds upon thin layer chromatography (tlc). An nmr examination of 15 in a second solvent produced an identical spectrum following which the unchanged compound was recovered. This result seems to eliminate the possibility that a single isomer of 15 is converted into two isomers by ring opening-reclosing in the nmr solvent ~ y s t e m . ~ In the tetramethyl derivative 14, as expected, the AMX system was replaced by an ABCD multiplet representing the CHzCHzgroups at positions 5 and 6 .
Experimental Section All melting points are corrected. Nmr spectra were obtained with Varian A-60 and HA-100 instruments employing tetramethylsilane as the internal standard. An Aerograph Model 202B dual-column gas chromatograph was used for vpc analysis. 3,3-Di(2-propynyl)-2,4-pentanedione (2),l 3-allyl-3-(2-propynyl)-2,4pentanedi0ne,~ 3,3-diallyl-2,4-pentanedione (6),6 and 3,3-dimethyl-2,4-pentanedione7were prepared as previously described a-(Hydrazinomethy1)benzyl Alcohol (l).-To a stirred solution of 186.9 g (5.84 mol) of hydrazine in 500 ml of butanol, held a t 86-92' by a steam cone, was added 60 g (0.5 mol) of styrene oxide over a 45-min period. After continued heating and stirring for 17 hr, 150 ml of water was carefully added and volatiles removed under reduced pressure. Distillation gave 54.0 g (71.0%) of product, bp 162-165' (4 mm) [lit.O165' (4 mm)] l-Glycidyl-4-methylpiperazine.-Was prepared from epichlorhydrin and 1-methylpiperazine by a method similar to that of g6
.
(4) A possible rationale of the melting point behavior suggested by a ref-
eree. (5) (8) (7) (8) (9)
E. K. Scbulte, J. Reiaoh, and A. Mock, Arch. Pharm., 196, 827 (1982). R. B. Davis and P. IIurd, J. Amev. Chem. Soc., 77, 3284 (1955). J. J. Bloomfield, J. Or#. Chem.. 16,4112 (1981). A. W. Johnson, E. Markham, and R. Price, O w . Syn., I S , 76 (1982). G. Benoit, Bull. SOC.Chim. Fr., 6 , 708 (1939).
CeH,C